nerdd-module 0.3.13__tar.gz → 0.3.15__tar.gz

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nerdd-module
-Version: 0.3.13
+Version: 0.3.15
 Summary: Base package to create NERDD modules
 Author-email: Steffen Hirte <steffen.hirte@univie.ac.at>
 Maintainer-email: Steffen Hirte <steffen.hirte@univie.ac.at>

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/config/models.py

@@ -71,12 +71,14 @@ class ResultProperty(BaseModel):
     name: str
     type: str
     visible_name: Optional[str] = None
+    visible: bool = True
     help_text: Optional[str] = None
     sortable: bool = False
     group: Optional[str] = None
     level: Level = "molecule"
     formats: Union[FormatSpec, IncludeExcludeFormatSpec, None] = None
     representation: Optional[str] = None
+    from_property: Optional[str] = None
     image_width: Optional[int] = None
     image_height: Optional[int] = None
 

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/converters/__init__.py

@@ -1,7 +1,8 @@
+from .basic_type_converter import *
 from .converter import *
 from .converter_config import *
-from .identity_converter import *
 from .mol_converter import *
 from .mol_to_image_converter import *
 from .problem_list_converter import *
+from .representation_converter import *
 from .void_converter import *

nerdd_module-0.3.13/nerdd_module/converters/identity_converter.py → nerdd_module-0.3.15/nerdd_module/converters/basic_type_converter.py

@@ -3,9 +3,9 @@ from typing import Any
 from .converter import Converter
 from .converter_config import ALL, ConverterConfig
 
-__all__ = ["IdentityConverter", "primitive_data_types"]
+__all__ = ["BasicTypeConverter", "basic_data_types"]
 
-primitive_data_types = [
+basic_data_types = [
     "int",
     "float",
     "string",
@@ -13,11 +13,11 @@ primitive_data_types = [
 ]
 
 
-class IdentityConverter(Converter):
+class BasicTypeConverter(Converter):
     def _convert(self, input: Any, context: dict) -> Any:
         return input
 
     config = ConverterConfig(
-        data_types=primitive_data_types,
+        data_types=basic_data_types,
         output_formats=ALL,
     )

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/converters/converter.py

@@ -25,7 +25,7 @@ class Converter(ABC):
 
     def __init__(self, result_property: ResultProperty, output_format: str, **kwargs: Any) -> None:
         super().__init__()
-        self.property = result_property
+        self.result_property = result_property
         self.output_format = output_format
 
     @classmethod

nerdd_module-0.3.15/nerdd_module/converters/mol_converter.py

@@ -0,0 +1,24 @@
+from typing import Any
+
+from ..config import ResultProperty
+from .converter import Converter
+from .converter_config import ConverterConfig
+
+__all__ = ["MolConverter"]
+
+
+class MolConverter(Converter):
+    def __init__(self, result_property: ResultProperty, output_format: str, **kwargs: Any) -> None:
+        super().__init__(result_property, output_format, **kwargs)
+
+    def _convert(self, input: Any, context: dict) -> Any:
+        if self.output_format == "sdf" and self.result_property.name != "input_mol":
+            # in an SDF, the main molecule (input_mol) can be a Mol object
+            return Converter.HIDE
+        elif self.output_format in ["sdf", "pandas", "record_list", "iterator"]:
+            return input
+
+    config = ConverterConfig(
+        data_types="mol",
+        output_formats=["sdf", "pandas", "record_list", "iterator"],
+    )

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/converters/mol_to_image_converter.py

@@ -9,14 +9,14 @@ from .converter_config import ConverterConfig
 
 __all__ = ["MolToImageConverter"]
 
-default_width = 400
-default_height = 300
+default_width = 300
+default_height = 180
 
 
 class MolToImageConverter(Converter):
     def _convert(self, input: Any, context: dict) -> Any:
-        width = self.property.image_width
-        height = self.property.image_height
+        width = self.result_property.image_width
+        height = self.result_property.image_height
 
         if width is None:
             width = default_width

nerdd_module-0.3.15/nerdd_module/converters/representation_converter.py

@@ -0,0 +1,42 @@
+from typing import Any
+
+from rdkit.Chem import MolToInchi, MolToSmiles
+
+from ..config import ResultProperty
+from .converter import Converter
+from .converter_config import ALL, ConverterConfig
+
+__all__ = ["RepresentationConverter"]
+
+
+class RepresentationConverter(Converter):
+    def __init__(self, result_property: ResultProperty, output_format: str, **kwargs: Any) -> None:
+        super().__init__(result_property, output_format, **kwargs)
+
+        representation = result_property.representation or "smiles"
+        if representation == "inchi":
+            self._serialize = MolToInchi
+        elif representation == "smiles":
+            self._serialize = MolToSmiles
+        else:
+            raise ValueError(f"Unsupported representation: {representation}")
+
+    def _convert(self, input: Any, context: dict) -> Any:
+        from_property = self.result_property.from_property
+
+        if from_property is None:
+            actual_input = input
+        else:
+            actual_input = context[from_property]
+
+        try:
+            representation = self._serialize(actual_input)
+        except:  # noqa: E722 (allow bare except, because RDKit is unpredictable)
+            representation = None
+
+        return representation
+
+    config = ConverterConfig(
+        data_types="representation",
+        output_formats=ALL,
+    )

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/input/inchi_reader.py

@@ -2,8 +2,8 @@ from codecs import getreader
 from typing import Any, Iterator
 
 from rdkit.Chem import MolFromInchi
-from rdkit.rdBase import BlockLogs
 
+from ..polyfills import BlockLogs
 from ..problem import Problem
 from .reader import ExploreCallable, MoleculeEntry, Reader
 from .reader_config import ReaderConfig
@@ -47,7 +47,7 @@ class InchiReader(Reader):
                     errors = []
 
                 yield MoleculeEntry(
-                    raw_input=line,
+                    raw_input=line.strip("\n"),
                     input_type="inchi",
                     source=("raw_input",),
                     mol=mol,

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/input/sdf_reader.py

@@ -2,8 +2,8 @@ from codecs import getreader
 from typing import Any, Iterator
 
 from rdkit.Chem import MolFromMolBlock
-from rdkit.rdBase import BlockLogs
 
+from ..polyfills import BlockLogs
 from ..problem import Problem
 from .reader import ExploreCallable, MoleculeEntry, Reader
 
@@ -59,7 +59,7 @@ class SdfReader(Reader):
                         errors = []
 
                     yield MoleculeEntry(
-                        raw_input=mol_block,
+                        raw_input=mol_block.strip("\n"),
                         input_type="mol_block",
                         source=("raw_input",),
                         mol=mol,

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/input/smiles_reader.py

@@ -2,8 +2,8 @@ from codecs import getreader
 from typing import Any, Iterator
 
 from rdkit.Chem import MolFromSmiles
-from rdkit.rdBase import BlockLogs
 
+from ..polyfills import BlockLogs
 from ..problem import Problem
 from .reader import ExploreCallable, MoleculeEntry, Reader
 from .reader_config import ReaderConfig
@@ -57,7 +57,7 @@ class SmilesReader(Reader):
                     errors = []
 
                 yield MoleculeEntry(
-                    raw_input=line,
+                    raw_input=line.strip("\n"),
                     input_type="smiles",
                     source=("raw_input",),
                     mol=mol,

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/model/convert_representations_step.py

@@ -12,13 +12,17 @@ class ConvertRepresentationsStep(MapStep):
         self, result_properties: List[ResultProperty], output_format: str, **kwargs: Any
     ) -> None:
         super().__init__()
+        self._result_properties = result_properties
         self._converter_map = {
             p.name: Converter.get_converter(p, output_format, **kwargs) for p in result_properties
         }
 
     def _process(self, record: dict) -> dict:
         result = {
-            k: self._converter_map[k].convert(input=v, context=record) for k, v in record.items()
+            k.name: self._converter_map[k.name].convert(
+                input=record.get(k.name, None), context=record
+            )
+            for k in self._result_properties
         }
 
         return {k: v for k, v in result.items() if v is not Converter.HIDE}

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/model/simple_model.py

@@ -113,26 +113,60 @@ class SimpleModel(Model):
         task_based_property = []
         if task == "atom_property_prediction":
             task_based_property = [
-                {"name": "atom_id", "type": "int"},
+                {"name": "atom_id", "type": "int", "visible": False},
             ]
         elif task == "derivative_property_prediction":
             task_based_property = [
-                {"name": "derivative_id", "type": "int"},
+                {"name": "derivative_id", "type": "int", "visible": False},
             ]
 
         default_properties_start = [
-            {"name": "mol_id", "type": "int"},
+            {"name": "mol_id", "type": "int", "visible": False},
             *task_based_property,
-            {"name": "input_text", "visible_name": "Input text", "type": "string"},
-            {"name": "input_type", "visible_name": "Input type", "type": "string"},
-            {"name": "source", "visible_name": "Source", "type": "string"},
+            {
+                "name": "input_text",
+                "visible_name": "Input text",
+                "type": "string",
+                "visible": False,
+            },
+            {
+                "name": "input_type",
+                "visible_name": "Input type",
+                "type": "string",
+                "visible": False,
+            },
+            {
+                "name": "source",
+                "visible_name": "Source",
+                "type": "string",
+                "visible": False,
+            },
             {"name": "name", "visible_name": "Name", "type": "string"},
-            {"name": "input_mol", "visible_name": "Input SMILES", "type": "mol"},
+            {
+                "name": "input_mol",
+                "visible_name": "Input Structure",
+                "type": "mol",
+                "visible": False,
+            },
+            {
+                "name": "input_smiles",
+                "visible_name": "Input SMILES",
+                "type": "representation",
+                "from_property": "input_mol",
+                "visible": False,
+            },
             {
                 "name": "preprocessed_mol",
-                "visible_name": "Preprocessed SMILES",
+                "visible_name": "Preprocessed Structure",
                 "type": "mol",
             },
+            {
+                "name": "preprocessed_smiles",
+                "visible_name": "Preprocessed SMILES",
+                "type": "representation",
+                "from_property": "preprocessed_mol",
+                "visible": False,
+            },
         ]
 
         default_properties_end = [

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/output/sdf_writer.py

@@ -1,6 +1,6 @@
 from typing import IO, Any, Dict, Iterable
 
-from rdkit.Chem import SDWriter
+from rdkit.Chem import Mol, SDWriter
 
 from .file_writer import FileLike, FileWriter
 
@@ -18,13 +18,22 @@ class SdfWriter(FileWriter, output_format="sdf"):
                 # assume that there is a mol object
                 mol = entry["input_mol"]
 
+                # if the molecule is erroneous, use an empty molecule
+                if mol is None:
+                    mol = Mol()
+
                 # write (almost) all properties to the mol object
                 for key, value in entry.items():
-                    value_as_str = str(value)
-                    if "\n" in value_as_str:
-                        # SDF can't write multi-line strings
+                    # skip "input_mol" key, because we use it as the main molecule
+                    if key == "input_mol":
                         continue
 
+                    value_as_str = str(value)
+
+                    # SDF can't write multi-line strings
+                    # -> replace newline with space
+                    value_as_str = value_as_str.replace("\n", " ")
+
                     mol.SetProp(key, value_as_str)
 
                 # write molecule

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/polyfills/__init__.py

@@ -1,3 +1,4 @@
+from .block_logs import *
 from .files import *
 from .get_entry_points import *
 from .literal import *

nerdd_module-0.3.15/nerdd_module/polyfills/block_logs.py

@@ -0,0 +1,45 @@
+from typing import Optional, Protocol, Type
+
+from rdkit.rdBase import BlockLogs as OriginalBlockLogs
+
+__all__ = ["BlockLogs"]
+
+
+class BlockLogsProtocol(Protocol):
+    def __init__(self) -> None: ...
+    def __enter__(self) -> "BlockLogsProtocol": ...
+
+    def __exit__(
+        self,
+        exc_type: Optional[Type[BaseException]],
+        exc_val: Optional[BaseException],
+        exc_tb: Optional[Type[BaseException]],
+    ) -> None: ...
+
+
+BlockLogs: Type[BlockLogsProtocol]
+
+if hasattr(OriginalBlockLogs, "__enter__"):
+    BlockLogs = OriginalBlockLogs
+else:
+    # We would like to use
+    #   with BlockLogs(): ...
+    # but this does not work with old versions of RDKit. Therefore, we create an instance of
+    # BlockLogs that will suppress log messages as long as it exists. When it is deleted (in the
+    # "__exit__" block), logs are enabled again.
+    class CustomBlockLogs:
+        block_logs: Optional[OriginalBlockLogs] = None
+
+        def __enter__(self) -> BlockLogsProtocol:
+            self.block_logs = OriginalBlockLogs()
+            return self
+
+        def __exit__(
+            self,
+            exc_type: Optional[Type[BaseException]],
+            exc_val: Optional[BaseException],
+            exc_tb: Optional[Type[BaseException]],
+        ) -> None:
+            del self.block_logs
+
+    BlockLogs = CustomBlockLogs

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/preprocessing/chembl_structure_pipeline.py

@@ -2,8 +2,8 @@ import warnings
 from typing import List, Optional, Tuple
 
 from rdkit.Chem import Mol
-from rdkit.rdBase import BlockLogs
 
+from ..polyfills import BlockLogs
 from ..problem import Problem
 from .preprocessing_step import PreprocessingStep
 
@@ -15,22 +15,16 @@ warnings.filterwarnings(
 )
 
 # We check if chembl_structure_pipeline is installed. Since importing this library already logs
-# messages, we suppress them using RDKit's BlockLogs. We would like to use
-#   with BlockLogs(): ...
-# but this does not work with old versions of RDKit. Therefore, we create an instance of
-# BlockLogs that will suppress log messages as long as it exists. When it is deleted (in the
-# "finally" block), logs are enabled again.
-block_logs = BlockLogs()
-try:
-    from chembl_structure_pipeline import get_parent_mol, standardize_mol
-
-    import_error = None
-except ImportError as e:
-    # raise ImportError later when using this class
-    # --> this allows to use the rest of the package without chembl_structure_pipeline
-    import_error = e
-finally:
-    del block_logs
+# messages, we suppress them using RDKit's BlockLogs.
+with BlockLogs():
+    try:
+        from chembl_structure_pipeline import get_parent_mol, standardize_mol
+
+        import_error = None
+    except ImportError as e:
+        # raise ImportError later when using this class
+        # --> this allows to use the rest of the package without chembl_structure_pipeline
+        import_error = e
 
 __all__ = ["GetParentMolWithCsp", "StandardizeWithCsp"]
 

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/tests/models/AtomicMassModel.py

@@ -64,3 +64,5 @@ class AtomicMassModel(SimpleModel):
                 {"name": "mass", "type": "float", "level": "atom"},
             ],
         }
+
+

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/tests/models/MolWeightModel.py

@@ -1,6 +1,7 @@
+from rdkit.Chem.rdMolDescriptors import CalcExactMolWt
+
 from nerdd_module import SimpleModel
 from nerdd_module.preprocessing import Sanitize
-from rdkit.Chem.rdMolDescriptors import CalcExactMolWt
 
 __all__ = ["MolWeightModel"]
 
@@ -47,4 +48,4 @@ class MolWeightModel(SimpleModel):
             "result_properties": [
                 {"name": "weight", "type": "float"},
             ],
-        }
+        }

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module/tests/representations.py

@@ -2,12 +2,11 @@ import numpy as np
 from hypothesis import given as hgiven
 from hypothesis import seed, settings
 from hypothesis import strategies as st
-
 # TODO: remove "type: ignore" later
 from hypothesis_rdkit import mols  # type: ignore
 from pytest_bdd import given, parsers
 from rdkit.Chem import MolToInchi, MolToMolBlock, MolToSmiles
-from rdkit.rdBase import BlockLogs
+from ..polyfills import BlockLogs
 
 
 @given(parsers.parse("a random seed set to {seed:d}"), target_fixture="random_seed")
@@ -69,6 +68,11 @@ def molecules_with_none(num, num_none=None, random_seed=0):
 
     generate()
 
+    for m in result:
+        if m is None:
+            continue
+        m.SetProp("_Name", "mol")
+
     # replace random entries with None
     indices = np.random.choice(num, num_none, replace=False)
     for i in indices:

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nerdd-module
-Version: 0.3.13
+Version: 0.3.15
 Summary: Base package to create NERDD modules
 Author-email: Steffen Hirte <steffen.hirte@univie.ac.at>
 Maintainer-email: Steffen Hirte <steffen.hirte@univie.ac.at>

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/nerdd_module.egg-info/SOURCES.txt

@@ -21,12 +21,13 @@ nerdd_module/config/package_configuration.py
 nerdd_module/config/search_yaml_configuration.py
 nerdd_module/config/yaml_configuration.py
 nerdd_module/converters/__init__.py
+nerdd_module/converters/basic_type_converter.py
 nerdd_module/converters/converter.py
 nerdd_module/converters/converter_config.py
-nerdd_module/converters/identity_converter.py
 nerdd_module/converters/mol_converter.py
 nerdd_module/converters/mol_to_image_converter.py
 nerdd_module/converters/problem_list_converter.py
+nerdd_module/converters/representation_converter.py
 nerdd_module/converters/void_converter.py
 nerdd_module/input/__init__.py
 nerdd_module/input/depth_first_explorer.py
@@ -61,6 +62,7 @@ nerdd_module/output/record_list_writer.py
 nerdd_module/output/sdf_writer.py
 nerdd_module/output/writer.py
 nerdd_module/polyfills/__init__.py
+nerdd_module/polyfills/block_logs.py
 nerdd_module/polyfills/files.py
 nerdd_module/polyfills/get_entry_points.py
 nerdd_module/polyfills/literal.py

{nerdd_module-0.3.13 → nerdd_module-0.3.15}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "nerdd-module"
-version = "0.3.13"
+version = "0.3.15"
 description = "Base package to create NERDD modules"
 readme = "README.md"
 license = { file = "LICENSE" }

nerdd_module-0.3.13/nerdd_module/converters/mol_converter.py

@@ -1,41 +0,0 @@
-from typing import Any
-
-from rdkit.Chem import MolToInchi, MolToSmiles
-
-from ..config import ResultProperty
-from .converter import Converter
-from .converter_config import ALL, ConverterConfig
-
-__all__ = ["MolConverter"]
-
-
-class MolConverter(Converter):
-    def __init__(self, result_property: ResultProperty, output_format: str, **kwargs: Any) -> None:
-        super().__init__(result_property, output_format, **kwargs)
-
-        if output_format == "sdf" and result_property.name == "input_mol":
-            # in an SDF, the main molecule (input_mol) can be a Mol object
-            self._serialize = lambda x: x
-        elif output_format in ["pandas", "record_list", "iterator"]:
-            self._serialize = lambda mol: mol
-        else:
-            representation = result_property.representation or "smiles"
-            if representation == "inchi":
-                self._serialize = MolToInchi
-            elif representation == "smiles":
-                self._serialize = MolToSmiles
-            else:
-                raise ValueError(f"Unsupported representation: {representation}")
-
-    def _convert(self, input: Any, context: dict) -> Any:
-        try:
-            representation = self._serialize(input)
-        except:  # noqa: E722 (allow bare except, because RDKit is unpredictable)
-            representation = None
-
-        return representation
-
-    config = ConverterConfig(
-        data_types="mol",
-        output_formats=ALL,
-    )